Our objective is to define molecular and biophysical mechanisms of learning and memory. Emphasis is placed on learning and memory which can be related to human cognition. Ultimate goals are to arrive at clinically meaningful interventions, and to design and construct artificial intelligence which has advanced learning and memory capabilities. With human cognitive functions as the principle frame of reference, the research focuses on associative processes (such as Pavlovian conditioning) rather than non-associative behavioral modifications (such as sensory adaptation, habituation, arousal and sensitization). The biological basis of learning and memory is of interest at several levels of complexity: behavioral phenomena, neuronal systems, neuronal membranes, and molecular transformations. To literally reconstruct the physiology involved (and then to model it in artificial contexts) it is necessary to use both "simple system" preparations such as the nudibranch mollusc Hermissenda crassicornis as well as "complex system" preparations such as rabbits. The molluscan work thus far has yielded the first unequivocal biological record of an associative memory. This record consists of persistent transformations of specific ionic channels. Because these records have been found within the membranes of identified single neurons it has proven possible to define biochemical pathways which regulate such long-term membrane modifications as well as to analyze how this biological memory record is expressed by the integrative functions of an entire neuronal system. The work on the vertebrate brain offers two essential opportunities. First, the generality of mechanisms determined for much less evolved species can be tested. Remarkably, the same ionic channel transformations have been shown in our program to record associative memory in the rabbit as were found in Hermissenda. Rabbit neural systems have also provided sufficient quantities of tissue so that conditioning-specific alterations of critical enzymatic (e.g., C-kinase) pathways which control membrane excitability have recently been demonstrated.